It was suggested by Ronald Fisher in 1931 that genes that benefit the male (including those required for spermatogenesis) would accumulate on the Y chromosome. Following the discovery that microdeletions of the Y chromosome were associated with diverse spermatogenic phenotypes, at least three intervals that contain one or more genes controlling male germ-cell differentiation have been identified in humans. These intervals, named AZFa, AZFb and AZFc, have been mapped, cloned and examined in detail for the presence of functional genes. In this review, I have discussed the genes that map to the AZF intervals and the evidence indicating which ones are the most likely candidates underlying Y-linked male infertility. In addition, I have considered the analysis of key intervals on the mouse Y chromosome, where it provides comparative data supporting the role of a candidate gene in an infertility phenotype.